Conventional cancer radioimmunotherapy, using antibody-radioisotope conjugates, has not demonstrated efficacy in clinical trials. It has not been possible to consistently deliver a sufficiently high radiation dose to tumor because of dose limiting radiation toxicity to bone marrow. Pretargeting a bispecific antibody to tumor, allowing non-target antibody to clear from circulation and then injecting the therapy radioisotope on a small molecular weight, bivalent hapten reactive with the bispecific antibody overcomes this limitation. The bivalent hapten cross-links tumor bound bispecific antibody enhancing its retention at the tumor. The bispecific antibody humanized MN14 (anti-CEA) x m734 (anti-In-DTPA) will be constructed by bis-maleimide cross-linking of the respective Fab' fragments. The construct will be physically and functionally characterized for monomericity and integrity of both binding activities on the same molecule. A series of animal studies in tumored mice is planned to determine flu in vivo performance characteristics of the bispecific antibody and the bivalent hapten (In-DTPA) Tyr (131I) Lys (In-DTPA). At the conclusion of these studies a bispecific antibody suitable for human clinical trials will have been constructed, characterized, and its function within the pretargeting protocol together with bivalent radio-hapten will have been defined. These data will be used in planning Phase I clinical studies. PROPOSED COMMERCIAL APPLICATIONS: The bispecific antibody and bivalent hapten pair made and characterized as described in this application are final formulations for committed clinical development as cancer therapy. The anti-CEA specificity of the antibody will allow potential therapy for most of the major solid tumor types including lung and colon cancers, our first targets. The beta-emitting therapy radioisotope 131I is readily available and relatively inexpensive.